What question did this study set out to answer?

This research aims to improve early identification of U.S. recession periods using machine learning models and macro-financial indicators.

Machine Learning Models for Early Warning of Financial Crises in the U.S. Economy Using Macro-Financial Indicators

View Full Paper

Ask AI

Bookmark

Authors

Models predict recession onset in the U.S. economy, suggesting better warning signals for policymakers.

This research aims to improve early identification of U.S. recession periods using machine learning models and macro-financial indicators.
Utilized logistic regression, random forest, and XGBoost models for prediction.
Analyzed a monthly dataset from 1970 to 2025, defining recession periods using NBER dates.
Conducted binary classification for recession prediction with time-series cross-validation.
Evaluated model performance using ROC-AUC, precision-recall curves, Brier scores, and calibration curves.
XGBoost and random forest models outperformed logistic regression by 5–10% on ROC-AUC.
The yield curve spread, unemployment, and VIX are the most reliable indicators for recession signaling.
Models provide warnings about 5–6 months before recession starts, with accuracy decreasing during economic shifts.

Rahman et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark